Thursday, August 13, 2009

The Blackberry and the pavement.

I recently learned an important lesson about Blackberry cell phones. They don't necessarily survive an impact with the roadway at 50 miles per hour. Of course I suppose I wouldn't expect them to.

"How did I discover this?" you may ask. I have the wonderful gift of being forgetful, and I left my Blackberry on top of the car while driving to work. I was lucky that I at least heard it falling off of the car, just to see it skip off the pavement a few times before careening off into the grass on the side of the road.

I turned around, not actually sure at the time what had fallen, but on the way back, I checked my pocket and figured it out. As I walked up to the phone, I noticed it had come out of it's case, but it looked intact. The case was lying about 10 feet away. When I first picked it up, I initially thought it was okay, minus the road rash. I later discovered otherwise.

Everything seemed to work, but it was obvious that the factory headphones which had been plugged in to it were garbage. The phone probably would have been alright if not for that, because the headphone connector was pretty well messed up. The plastic housing had cracked and one or more of the contacts on the board had possibly come loose. Now the phone doesn't seem to be able to tell if headphones are plugged in or not - it keeps switching back and forth from the internal speaker to the headphones.

I decided I might be able to fix it at least well enough to get it to work, so I bought a T5 screwdriver from Menards and went to work on it. After completely disassembling the unit and closely inspecting the jack, I couldn't figure out what was wrong with it. I assumed some of the contacts were shorted due to the shattered plastic housing, but I couldn't tell where. After a few hours of probing, re-assembling and disassembling again to no avail, I decided it was time to just pop the headphone connector off of the board. Surely that would fix any shorts in the connector and at least allow me to use the phone.

Wrong. The phone still can't seem to determine if headphones are connected or not. I even tried using a bluetooth headset, but the uncertainty about the headphones causes the phone to continuously try to switch inputs. Right now I'm of the opinion that of the 6 contacts inside the headphone connector, there must be some that are normally closed when nothing is connected, but I haven't been able to determine which ones yet, as the board contacts are very difficult to get to when the phone is assembled enough to power on. Eventually I will probably solder some leads onto the board to allow me to do some further testing.

For now, I've found someone selling an unlocked T-mobile G1. I've wanted a G1 for a while now anyway. I met with him last night, but we couldn't get it to work with AT&T. After some research, apparently there was an APN setting that needed to be changed to use the AT&T data network, because he apparently has it working now with an AT&T SIM, so we'll try again soon I hope.

Friday, July 10, 2009

The Toshiba NB205 - part 2

So it seems there has at least been a partial solution to the sound problem. Putting the line:
options snd-hda-intel model=asus-mode4

into /etc/modprobe.d/alsa-base.conf gives sound through the headphones, although the master volume control doesn't work, and there is no sound from the built-in speaker. I am satisfied for now, and I'll post any solutions that make this work better. Thanks to yorkzhang for the solution to this and for finding me on github to give it to me.

Tuesday, June 23, 2009

The Toshiba NB205

I recently bought my first netbook, the Toshiba NB205. I saw it at the Buy More on display for $399, so as I always do with netbooks on display, I tested my typing on the keyboard. Contrary to most netbooks, I could actually type on this one relatively easily. So I did some research to see if there were any problems getting Linux to run on it. I found nothing, presumably because this was Toshiba's first netbook in the US, and it had just been released. So I looked at the specs, saw mostly Intel chipsets, which usually work well in Linux, and decided to buy it.

My first task was to image the drive onto my external HD so I could put it back to the way I bought it if the need ever arose. I didn't intend to use the Windows XP that was on it, but I wanted to have an image "just in case". This turned out to be a good move, which I'll explain later.

I went with Ubuntu Netbook Remix for the laptop, because Ubuntu usually works well for me and it seemed like the logical option. Upon loading the image onto my thumb drive, I noticed the wireless didn't work. lspci revealed to me that it had an Atheros AR9285 wifi chipset, not Intel as I had expected. Luckily Atheros decided to release an open source driver for this chip, but it was in the 2.6.29 kernel, not the 2.6.28 which was on the live image. I found that I would need to install linux-backports-modules-jaunty after install to get it working - no problem.

So I installed Ubuntu on the laptop, wiping XP. I ran apt-get install linux-backports-modules-jaunty, rebooted, and presto, it detected my wireless card. All was well, or so I thought. sudo iwlist wlan0 scan revealed no scan results, but I was 2 feet from my router (not to mention the 20 other networks I usually pick up in my neighboorhood). I noticed the wifi light on the front of the machine was not lit. The card must not be enabled. After scouring the internet for several hours and looking thoroughly through the bios, I found no way to enable the wireless card. I remembered a friend of mine once had a similar issue with a laptop - no way to enable the card in Linux if it had been turned off in Windows.

So I reconnected my external HD to the laptop and booted from the live usb image of Ubuntu and proceeded to write the drive image back onto the internal hard drive and left it for the night (It took a while over USB, the netbook didn't have any eSATA or firewire ports).

The next morning I found the image had been written back to the drive, so I rebooted and proceeded to boot Windows once to try to enable the wireless card. I tried pressing Fn+F8 on the setup screen, but nothing happened, so I continued with the initial setup. After Windows booted fully, I pressed it again and finally, the wifi light lit up. I shut down Windows and booted from the thumb drive again.

When it booted up, I noticed that the wifi light was still on. Good. This time I opted to leave Windows XP, with as little space as was practical, and set up dual boot, just in case I needed it again. After installing Linux again and installing the backports, I tried a scan again and my network showed up. The Gnome network manager worked well with my wpa encryption and all was well for the moment.

After using the laptop for about a day I noticed I hadn't heard any sounds, so I tried to play an mp3 file. Sure enough, nothing. I checked all the mixer settings and played with them a bit, but nothing seemed to work. So again here I go scouring the internet for solutions. After a few hours of searching I finally came across the alsa-project bug tracker and found bug #4575, which seemed to be the same problem I have. Same sound chip (Intel 82801G with Realtek ALC272) and no sound, so I'm assuming this is an alsa bug. For now I guess I'll do without sound, but I hope it's fixed soon.

Overall this netbook is really nice. The keyboard still feels really good and is easy to type on. It also has great battery life, about 6-7 hours with wifi on. UNR runs very well on it and I haven't had any problems with suspend-resume yet, which is crucial for a netbook. It's more than I can say with my Dell Latitude with Kubuntu, where resuming is about a 50/50 proposition. I beleive I'll be very happy with this netbook once the sound is working, but for now I'm satisfied with my purchase.